MedCase: a template medical case store for case-based reasoning in medical decision support

• Authors: Hsien-Tseng Wang;Abdullah Uz Tansel
• Venue: Proceedings of the 2013 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining
• Year: 2013

The early development of medical decision support systems (appeared as expert systems (ES)) mainly focused on, among others, rule-based reasoning (RBR) and decision table/tree (DT) methods as problem solving strategies. These efforts were novel at the time; however, as these methodologies applied to more complex situations, the construction of knowledge base (e.g. rules, cases and 'models') for specific problem solving tasks becomes difficult and time consuming. Remedies to these difficulties have been sought, aiming at better knowledge modeling, knowledge acquisition, and extending the problem solving paradigm to distributed architectures. Alternatively, case-based reasoning (CBR) provides a different problem solving paradigm. In CBR, the knowledge is seen as cases that contain explicit and implicit aspects of the knowledge for solving a problem. The CBR methodology works in a practical way, and the reasoning is based on recalled knowledge memory of solved cases. To alleviate the difficulty of knowledge (case) acquisition and construction, this paper presents a design of a template case store, called MedCase. MedCase utilizes the semantic web technologies and supports a distributed CBR system architecture. MedCase promotes an open and accessible architecture for common CBR tasks in a virtual Healthcare Enterprise environment. MedCase will also allow the construction and sharing of cases that facilitate the development of distributed CBR-based medical decision support systems.